Damping and H/V

[luigiV]

Hi Marc,
In a your work you used damping to prove that the H/V peak was natural. This is interesting because it adds to the "classic" satisfaction of the SESAME parameters to declare naturalness.
In the Geopsy documentation there are indications for setting the parameters (size of the window and suggestion of the order of the Butterworth filter), but the reality seems more complex to me. In fact, the value of z for the same data set appears directly proportional to the bandwidth of the bp filter and instead inversely proportional to the order of the Butterworth filter. Thus, when the z value is close to the 5% threshold, it would be enough to slightly modify one of these parameters to reach the goal of a z>5% value and thus "force" the conclusion of naturalness. Instead, using a frequency dependent filter I observed that the damping is almost always <5% whatever the window tape adopted.
So I ask if there are some criteria for a correct setting of the damping toolbox parameters.

Thanks
Luigi

[admin]

Effectively, with a very narrow bandwidth, the attenuation is always lower than with larger one. This is the heart of Fourier transforms. On the extreme side, if the band is so small that only one frequency remains after filtering, the whole signal reduces to a single cosine which damping is null.

Natural or industrial should not been seen as white or black. If you have a reasonable bandwidth, using a Butterworth filter, and if the attenuation is not visible over the first few arches (probably an attenuation close to or lower than 5%), there are suspicions of a sustain (industrial) peak. If you target is soil characterization, then do not interpret that peak. If your target is the identification of that peak, do eventually array measurements to localize it.

[luigiV]

Thanks a lot